Hi, I'm Dianne Nice, community editor for Report on Business, and I'll be hosting today's debate. I'd like to welcome Ben Rabidoux, an advisor and strategist at M Hanson, and Eric Lascelles, chief economist at RBC Global Asset Management.
Thank you, Dianne. And thank you to the Globe and Mail for highlighting this very relevant topic and to Mr. Lascelles for taking the time out of his no doubt busy schedule to engage in this important dialogue.
Thanks very much, great to be here.
Ben and Eric, you've spent the week debating the threat of Canada's household debt on our economy. It's been a fascinating discussion. Today is the final debate, in which readers will have a chance to question your positions.
Some readers have already submitted questions by email, so let's get started.
This first question comes from Tom Hobbs:
Statistics show that the biggest change in household debt is the size of the mortgage that a family now pays.
When a family, even with a flat income, downsizes their home the amount of disposable income increase significantly. They can afford to pay down debt and buy cars etc.
The poor recovery from the economic crisis can be stimulated by building lots of low cost housing across Canada and by motivating families to buy them.
I would like to see this specific issue debated with concrete proposals for how to make it happen. Personally I think the easiest way to do this is to focus on stimulus to build new low cost houses for seniors and to sub-divide existing large houses and lots.
The high cost of housing is hurting Canada's international competitive position.
I think this is a question you would both like to weigh in on.
Boy, that's a big comment/question. Let me break my response into two parts. First, how would you stimulate the production of smaller homes? I suppose you'd do it by offering tax incentives or subsidies to the builders or buyers. Or you could somehow swing a more favourable mortgage rate (or perhaps cheaper mortgage insurance) on an especially small mortgage... However, I'm just not convinced that's a likely or viable path...
First let me say that I do believe the current shift towards larger homes is likely to reverse course in the next few years as a simple function of demographics and affordability. So I do agree there.
I suppose there are a couple issues with the proposal. One is that in order to realize the cost benefits, the current, larger home must be sold at current valuations. I'm not sure those will exist indefinitely and I see the greatest risk of price compression in these larger homes.
The second issue is that by all measures, we have built more homes than demographics warrant over the past decade. The US peaked at 2.4 people per dwelling. We're now at 2.3. I'm not sure that adding new dwellings into an already-saturated market with record high ownership rates is the ultimate solution.
...The reason is just that Canadians are clearly demonstrating an appetite for fairly large homes. They're buying what they want. Builders are merely responding to demand. Now onto the second part of my response. I think you've made a useful point, which is that Canadians actually have a fair bit of wiggle room with their debt. If/when interest rates rise, I am somewhat assured that Canadians could indeed downsize their home, sell their second car, scale back on their RRSP contributions or make any number of other adjustments to live until another day.
Here's another question I think you can both respond to. It comes from Jay LaRochelle: How much does Canada's student debt (including all federal, provincial, and bank loans) contribute to the current household debt situation? Is student debt itself causing the need for Canadians to take on other household debt, just to be able to afford things like a car?
I don't have specific data at my fingertips. Perhaps Eric might. But I do agree that rising student indebtedness relative to starting wages is very problematic. In particular, I'd be concerned about what indebtedness in this demographic means for housing demand, consumption, and even household formation rates going forward. It certainly is a concern, but not one that I've dedicated a lot of time to trying to quantify.
I'll certainly defer to Ben on this in the event that he has some good statistics, but my understanding is that total student debt in Canada is fairly low -- around $15-20 billion out of total household debt of $1.6 trillion, so not a huge share. I can certainly appreciate that it is incredibly burdensome to many, despite not budging the needle at the aggregate level.
However, as someone who had an earlier career researching university education issues, let me say that the return on university education is still stunningly good for most students. When I think about debt, this is "good debt", ie debt that actually creates income down the line. Mortgage debt has a more neutral association -- at least it is going towards an asset -- whereas something like credit card debt is much more negative.
I agree with Eric's assessment here.
Ben and Eric, we often hear the argument that there are good and bad types of debt. What do you think?
Is Canadian household debt made up of a good mix of debts? What do the numbers tell us?
It's harder than I'd like to admit to distinguish between the two, but I think there's something to the idea of good/bad debt. One way would just be to think about whether you can afford your debt. If not, that's clearly "bad."
Thinking of it a different way, I laid out a simple framework in my last response. Student debt serves a very useful purpose and pays for itself over time, so is usually "good." Mortgage debt is in the middle -- you at least get a lasting asset for it, and you avoid having to pay monthly rent. Credit card debt is not ideal given high interest rates and the sorts of things that usually show up on credit card bills (usually not things that impart lasting benefit.)
More generally, I think we need to get away from thinking of debt as a purely negative thing. Families have low incomes and large expenses when they get going, and then large incomes and lower expenses later on. It makes perfect sense for the average household to borrow from the future to smooth this out.
Well there is certainly some truth to this. In most cases, debt to invest in your own earning power (i.e. education) can be an excellent and wise form of debt. Even mortgage debt, provided the buyer has stress-tested their finances and understands the risks to their equity by buying at this point in the credit and housing cycle, can be seen as good debt over the long term (emphasis on the LONG term). However, debt used to finance current consumption is problematic and can lead to significant imbalances in the broader economy when it happens on a large scale. Ultimately, even "good" debt can be an anchor in rough economic times. It should always be treated with a great deal of respect.
Ben I suspect you'll disagree given comments in one of your earlier editorials, but my feeling is that Canadian household debt has a pretty good composition. The significant majority is mortgage debt. As I say, it secures an asset, and avoids rental payments.
Eric, would you like to respond to the question about house prices?
I agree with Ben, home prices would be somewhat lower. As to whether they'd be "much" lower, I'm less sure. Most Canadians do not rely upon CMHC to finance their mortgages. Presumably, banks would either not lend to the highly leveraged group (in which case they'd be renting and somebody else would own the home), or banks would increase mortgage rates for this group to an appropriate level. That higher appropriate level wouldn't necessarily be much different than the insurance fee that people currently pay to CMHC. After all, CMHC has made money at that insurance fee, I don't see why banks wouldn't. Still, this is a very theoretical question -- the reality is that CMHC does exist and plays a significant role.
Eventually composition will take a back seat to total debt outstanding. Aggregate debt burdens must be serviced and repaid out of the aggregate wealth within an economy. You can't squeeze blood from a stone, and I'm more than a little nervous seeing these debt burdens reach 95% of GDP....or roughly where the US peaked. And I'd also point out that the percentage of household debt that was mortgage debt was much higher in the US than in Canada presently.
Again, I'd imaging Eric would have a better sense of what's happening on the front lines. On the "when" side, most of the B20 guideline changes are set to be implemented by October, but my understanding is that most major banks have already implemented a number of them: 65% LTV limits on HELOCs for example.
In response to YVRHOusing, I suspect you've given the question of B-20 OSFI guideline implementation more thought than me. I don't see anything in it that would suggest an aggressive change in the availability or pricing of mortgages, but there is no question there is a broader trend from the government in discouraging home equity lines of credit, and in tightening up rules. So we can assume that it will continue to get -- at the margin -- a bit more difficult to obtain a mortgage in Canada. And that may be a good thing -- having the appropriate rules is important.
As a rebuttal to one of Ben's earlier comments, I also get a bit nervous when Canadian household debt continues to rise relative to income, but keep in mind several things. First, Asset levels have also risen nicely. This would be true even if home prices had not gone up. Second, interest rates are low, meaning the debt feels fairly cheap. Third, we don't know what the "limit" for leverage is. It seems to go up over time. It seems to be higher in rich countries than poor countries. Canada is now higher than the U.S., but not nearly as high as Denmark or the Netherlands, which continue to plug along.
First of all, it is true on a national basis. This can be verified through Bank of Canada and US Fed data. Unfortunately we don't have city-specific debt figures, but I would imagine you're probably right that mortgage debt would be substantially higher in these areas.
It is tough to get good regional or micro data on housing and debt in Canada. We can talk about home prices in Vancouver relative to income. It is harder to talk about overall debt levels (though it is easy enough to guess that they would also be high). The work I've done on housing valuations in particular suggest that Toronto isn't especially out of line relative to the national affordability figures (ie higher home prices and higher incomes). However, Vancouver is. This is true on two fronts. First, Vancouverites have _always_ paid a larger share of their income towards mortgage payments. Second, their current payments are now running unusually high above the Vancouver norm. So the market is clearly more overdone there, and Vancouver households have more debt than the rest.
Inflated real estate prices are both a cause and a symptom of the current credit boom. As I noted in one of my editorials, 85% of household debt in Canada is tied to housing....either via mortgage debt or HELOC debt.
To Kent Chin, yes Denmark and Netherlands are the outliers. But the fact remains that they survive at that level. I just want to illustrate the point that there is nothing magical about a 150% household debt-to-income level. Others survive at much higher levels. The US and UK did not begin to delever because they hit a suddenly unsustainable level. They began to delever because their homes were foreclosed upon and their employment prospects were sufficiently damaged that they had to unload somes debt.
Also useful to recognize that debt-to-income ratios are comparing a stock to a flow. Would we feel better if instead of having a 150% household debt-to-income ratio, we called it a 4% debt-to-lifetime income ratio?
Real estate prices are definitely part of the story for the higher debt. I suppose I differ from Ben in believing that the conseqeuences of lower home prices / a weaker housing market / household deleveraging wont necessarily spell doom for the Canadian economy (slower growth, absolutely).
Here's a question from Aiden: Debt-to-income is often quoted, and is interesting to consider, especially the changes in the ratio, but it's not a good tool to determine over leverage. If someone has a mortgage and a good job, it's likely they will exceed this ratio without really having a problem.
How many Canadians are at or over their TDS ratio benchmarks, or amount of debt, property tax and home heating costs to income, and how much comfort do they have if rates go up? I.e. how far can interest rates go up before they breach their TDS thresholds? Then, how many have liquid reserves, and are they sufficient?
I agree with Eric that there is nothing "magical" about this level, but I do think the experiences of other countries, many of which began a painful debt deleveraging at lower debt-income figures than we currently hold, provide a useful warning sign. That we *can* stretch the debt and leverage band a little further before it snaps certainly doesn't mean we should.
To Aiden. Great comment. Debt-to-income is useful, but has its flaws. As I said in one of my editorials, I have no problem whatsover with a young couple earning $80,000 after taxes with a $300,000 mortgage. Yes, their debt ratio is over twice the national average. But they can probably afford it. This is why individual circumstances are so important, and a further reason why nationwide debt ratios don't have magic kill switches.
You correctly wonder what fraction of Canadians have too much debt. There isn't as much information as I'd like, but the Bank of Canada has done some good work on this. The answer is that well under 10% of Canadians are paying >40% of their income to service their debt. That's traditionally considered the danger zone. And in terms of actual mortgage delinquencies, this still runs well below 1% (US peaked at well north of 10%).
For sure, rising interest rates / a weaker economy would make that materially worse. But Canadians usually act quite responsibly. The microdata shows that seniors hold less debt than younger people, and the poor hold less debt than the rich. Most people scale their debt to what they can afford.
To Aiden's point: The Bank of Canada has done some stress testing, and we can get some useful insights on how vulnerable some households by looking at data compiled by CAAMP. The bottom line is that we are most concerned about those families at the bottom of the "debt pyramid", whose debt burdens are multiples of the "average".
The other point that I often discuss is that during a credit boom such as we are experiencing in Canada, the economy and labor market appears more resilient than it is absent the credit infusion. So I'm always a bit uneasy projecting current trends into the future when one trend is clearly unsustainable.
This question comes from Stella. It's tough to find this kind of breakdown. Can either of you comment?
It would be nice to get a breakdown on what Canadian household debt really entails. Mortgage debt is normal and should not be included unless it is affected by sub-prime mortgages. How much is due to car payments? Do Canadians pay off one car before buying another or add on the unpaid debt from the first? How much is due to a credit margin? How much to credit cards?
I agree with Ben about credit booms making the economy look better than it is. As a worst case scenario (and nowhere near my expectations for Canada), Ireland and Spain were both looking very good from an economic/government debt perspective until the housing market crumbled.
If you haven't seen our reader poll on household debt, check it out here. It's a fascinating look at how confident Globe readers are in their ability to pay off their own debts.
Do either of you, Ben and Eric, have a sense on what specific changes borrowers will see with the new B-20 OSFI guidelines implementation, and when? I've heard a lot about the new guidelines in a general sense but don't have a sense as to what specifically will change for people when they apply for loans, and what will change over the next while.
